Ventura N†, Rea S.L.†. Caenorhabditis Elegans

584_200600248_Ventura.qxd 30.04.2007 13:12 Uhr Seite 584

Biotechnology Journal DOI 10.1002/biot.200600248 Biotechnol. J. 2007, 2, 584–595

Research Article Caenorhabditis elegans mitochondrial mutants as an investigative tool to study human neurodegenerative diseases associated with mitochondrial dysfunction

Natascia Ventura1, 2* and Shane L. Rea1*

1Institute for Behavioral Genetics, University of Colorado at Boulder, Boulder, CO, USA 2Laboratory of Signal Transduction, Department of Experimental Medicine and Biochemical Sciences, University of Rome “Tor Vergata”, Rome, Italy

In humans, well over one hundred diseases have been linked to mitochondrial dysfunction and Received 14 December 2006 many of these are associated with neurodegeneration. At the root of most of these diseases lay in- Revised 5 March 2007 effectual energy production, caused either by direct or indirect disruption to components of the Accepted 13 March 2007 mitochondrial electron transport chain. It is surprising then to learn that, in the nematode Caenorhabditis elegans, a collection of mutants which share disruptions in some of the same genes that cause mitochondrial pathogenesis in humans are in fact long-lived. Recently, we resolved this paradox by showing that the C. elegans “Mit mutants” only exhibit life extension in a defined window of mitochondrial dysfunction. Similar to humans, when mitochondrial dysfunction becomes too severe these mutants also exhibit pathogenic life reduction. We have proposed that life extension in the Mit mutants occurs as a by-product of compensatory processes specifically activated to maintain mitochondrial function. We have also proposed that similar kinds of processes may act to delay the symptomatic appearance in many human mitochondrial-associated disorders. In the present report, we describe our progress in using the Mit mutants as an investigative tool to study some of the processes potentially employed by human cells to offset pathological mitochondrial dysfunction.

Keywords: Aging · Caenorhabditis elegans · Mitochondria · Mit mutants · Neurodegenerative disorder

1 Introduction a cause for numerous human diseases associated with neuronal degeneration [1]. Most of these mutations di- 1.1 Mitochondrial mutations associated with human rectly or indirectly affect mitochondrial electron transport neurodegenerative diseases chain (ETC) function. The subsequent alterations to the ETC have been shown to reduce ATP production, de- Over the last few decades, mitochondrial dysfunction, at- crease mitochondrial membrane potential, impair calci- tributable to mutations in either mitochondrial DNA um buffering, shift nucleotide pool ratios, increase reac- (mtDNA) or nuclear DNA (nDNA), has been established as tive oxygen species (ROS) generation, and ultimately result in cell death [2, 3]. Most of the neurodegenerative disorders associated with mitochondrial dysfunction display Correspondence: Dr. Natascia Ventura, University of Rome “Tor Vergata”, a late onset and a chronic progressive course. Typically, Via Montpellier 1, 00133 Rome, Italy symptoms associated with these diseases appear only af- E-mail: [email protected] ter mitochondrial dysfunction and neuronal degeneration Fax: +39-06-72596505 become very severe [4]. Abbreviations: ETC, electron transport chain; FA, Friedreich’s Ataxia; KO, knockout; mtDNA, mitochondrial DNA; ROS, reactive oxygen species;

UPRmito, unfolded protein response * Both these authors contributed equally to this work.

Biotechnol. J. 2007, 2, 584–595 www.biotechnology-journal.com

Friedreich’s Ataxia (FA) is the most commonly inher- eases [12]. Despite these considerations, recent advances ited ataxia and it is caused by pathologically low levels of in the nematode aging field have paradoxically revealed frataxin, a nuclear-encoded protein involved in mitochon- that mutations in many mitochondrial genes (collectively drial iron storage and Fe-S cluster biosynthesis [5, 6]. Sev- referred to as “Mit” mutations, and mainly affecting the eral components of the mitochondrial ETC contain Fe-S ETC), can actually prolong the lifespan in this organism clusters and hence require frataxin for their function. [13–15] (Table 1). Chronic frataxin deficiency induces a vicious cycle that We recently resolved this human-worm lifespan para- ultimately results in irreversible mitochondrial dysfunc- dox by showing that life extension in the C. elegans Mit tion and the complete dismantling of normal cell physiol- mutants only occurs within a limited window of mito- ogy. As for many mitochondrial-associated disorders, for chondrial dysfunction (Shane L. Rea, Nastascia Ventura FA, once a recognizable pathology presents, cell damage and Thomas E. Johnson, manuscript submitted). This life is usually so severe clinicians can only intervene with extension window was bounded by two thresholds: the symptomatic treatment. It would be highly preferable, first corresponded to moderate levels of mitochondrial then, if treatments could be established duting the dysfunction and was associated with slowed larval devel- presymptomatic phase of FA. By doing so, it may be pos- opment, small adult size, decreased fertility, and in- sible to delay the clinical appearance of this disorder, per- creased lifespan. When mitochondrial disruption became haps even preventing it permanently. too severe a second threshold became apparent – signifi- The Mitochondrial Threshold Effect Theory suggests cant larval arrest occurred, or adult animals were invari- that cells can cope with a certain degree of mitochondri- ably sterile, and lifespan was shortened. The C. elegans al dysfunction until a critical threshold is reached. Exper- Mit mutants, with their paradoxical mitochondrial defects imental evidence suggests that several factors underlie leading to lifespan extension, tell us that there can be an the mitochondrial threshold effect including, mtDNA het- advantage in lowering the mitochondrial functionality, eroplasmy, compensatory mitochondrial biochemical but only when it is kept within defined limits. pathways, and age-related mitochondrial changes [7]. We recently proposed that several, mutually nonexclusive, 1.3 C. elegans mitochondrial mutants as a model for stress response pathways might also act to help cells human neurodegenerative diseases associated with counter mitochondrial dysfunction. Such pathways could mitochondrial dysfunction be responsible for the chronic nature of neurodegeneration and the delayed appearance of many mitochondrial- We recently proposed that the kinds of compensatory associated disorders [8]. In support of this idea, it was re- mechanisms activated to counteract the mitochondrial cently shown that mice with moderate levels of frataxin dysfunction in mammalian cells are potentially the same inactivation have a significant alteration in their pattern as those activated in the C. elegans Mit mutants that ul- of expression of multiple genes; this was despite the fact timately lead to life extension. Similar compensatory that these animals manifested with no classical features mechanisms might also act to delay the appearance of mi- of neurodegeneration at neither the phenotypic nor bio- tochondrial-related disorders in humans [8]. We have also chemical level [9]. It is thus highly conceivable that com- proposed that the activation of such pathways represent pensatory pathways can be induced in cells with com- a continuum of responses and that the ability of each to promised mitochondrial function well before phenotypic effectively mask overt mitochondrial dysfunction de- markers of an established pathology appear. The genetic pends upon the severity of the lesion they are trying to nature of many mitochondrial diseases, coupled with the counter. If mitochondrial dysfunction becomes too severe presence of an asymptomatic window preceding the on- even these pathways are unable to compensate and we set of an established pathology, provides a great potential presume that this is the reason lifespan shortens in the for presymptomatic diagnosis and preventive therapy. Mit mutants and mitochondrial disorders suddenly appear in humans. As a first attempt to elucidate such com- 1.2 Mitochondrial mutations associated with longevity pensatory pathways, we recently generated a C. elegans in Caenorhabditis elegans model of FA using a bacterial feeding RNAi against the worm frataxin ortholog, frh-1. Consistent with other Mit The Free Radical Theory of Aging asserts that oxidative mutants, we found that mild reduction of frataxin expres- stress underlies the causes of aging [10]. Since mitochon- sion initially increased the animal lifespan [15]. More se- dria are generally believed to account for the majority of vere reduction, as expected, shortened the lifespan endogenous ROS production, they too have been impli- (Shane L. Rea, Nastascia Ventura and Thomas E. John- cated as one of the main culprits responsible for normal son, manuscript submitted). cellular aging [11]. Aging is one of the greatest risk factors We mentioned that several phenotypes coordinately behind several neurodegenerative disorders that have track with lifespan outcome in the C. elegans Mit mutants been associated with defective mitochondrial function, including altered larval development rate, reduced adult including Parkinson’s, Huntington’s, and Alzheimer’s dis- size, and diminished fertility (Table 1). The severity in the

Biotechnology Journal Biotechnol. J. 2007, 2, 584–595

Table 1. The Mit mutants of C. elegans (genetic and RNAi-defined) Mit phenotypesa) Human C. elegans gene Age Ste/Fer Emb HMADb) Gro Lva ortholog COMPLEX I (NADH:ubiquinone oxidoreductase) C09H10.3 (nuo-1, NDUFV1/51 kDa subunit) [54] [55] [56, 57] [55, 58] [56–58] NDUFV1 [59] T10E9.7 (nuo-2, NDUFS3/30 kDa subunit) [14, 60] [61] [57] – [57, 58, 61] NDUFS3 [23] Y57G11C.12 (nuo-3, NDUFA6/B14 subunit) [14, 60] [56] [57] – [56–58] NDUFA6 – K04G7.4 (nuo-4, NDUFA10/42 kDa subunit) [60, 62] [56] [57] [63] [56, 57] NDUFA10 – Y45G12B.1 (nuo-5, NDUFS1/75 kDa subunit) [60] [56] [57] – [56, 58] NDUFS1 [64] D2030.4 (NDUFB7/B18 subunit) [13, 62] [13, 61] [13, 57] – [57, 58] NDUFB7 [65] T20H4.5 (NDUFS8/23 kDa subunit) [62] [56] [57, 58] – [56, 57, 58] NDUFS8 [65] COMPLEX III (cytochrome c reductase) C54G4.8 (cyc-1, cytochrome c1) [14, 60] [61] [57] [58, 63] [57, 58, 61] CYC1 – F42G8.12 (isp-1, subunit RIP1, [27] – – [58, 66] [58, 66] UQCRFS1 – Rieske Fe-S protein) T02H6.11 (UQCRB) [13] [13] [13, 56, 57] [13, 58, 63] [5657, 58] UQCRB [67] COMPLEX IV (cytochrome c oxidase, COX) F26E4.9 (cco-1, subunit Vb/COX4) [13, 14, 60, 62] [13, 61] [57] [13, 58] [57, 58] COX5B – Y37D8A.14 (cco-2, subunit Va/COX6) [60] – [57] [56] [56–58, 63] COX5A – F26E4.6 (subunit VIIc/COX8) [13, 14] [13, 56] [13, 56] [13, 56, 58] [56, 58] COX7C – F54D8.2 (tag-174, subunit VIa/COX13) [62] ––––COX6A1 – W09C5.8 (subunit IV/COX5b) [13, 14] [13, 61] [57] [58, 13] [57, 58] COX4I2 – COMPLEX V (F1F0-ATP synthase) C34E10.6 (atp-2, beta subunit) [54] [55] [54, 55, 57, 63] [55, 56] [55, 57, 58] ATP5B – F27C1.7 (atp-3, subunit OSCP/ATP5) [14, 60] – [57, 61, 63] [57, 58] [57, 61] ATP5O – T05H4.12 (atp-4, subunit Cf6 [60] [56] [56] [56, 57, 63] [56, 58] ATP5J – (coupling factor 6)) C06H2.1 (atp-5, subunit d/ATP7) [60] [56, 63] [57] [63] [56–58] ATP5H – C53B7.4 (asg-2, subunit g/ATP20) [62] [55–57] – [55, 58] [55, 58] ATP5L – F02E8.1 (asb-2, subunit b/ATP4) [60] [55] [57] [58, 63] [57] ATP5F1 – Other ETC related F59G1.7 (frh-1, defective Fe-S [15] [56, 57] [15] [15] – FXN [68] containing enzymes) B0261.4 (mitochondrial S-ribosomal protein) [13] [13, 58, 61, 63] [57] [13, 57, 58] [58, 61] MRL47 – T06D8.6 (cchl-1, Holocytochrome [13] [13, 56, 57] [13] [56] [58, 69] CCHL [70] c synthase/heme-lyase) ZC395.2 (clk-1, DMQ mono-oxygenase/ [71, 72] [72] – [72] – COQ7 – ubiquinone biosynthesis protein COQ7/CLK-1/CAT5) ZC395.6 (gro-1,tRNA δ(2)- [73] [73] – [56, 58] – TRIT1 – isopentenylpyrophosphate transferase) ZK524.3 (lrs-2, leucyl-tRNA synthase) [13] [58, 63] – [69] [69] LARS2 [74, 75] T27E9.1 (tag-61, ANT2, ADP/ATP carrier) [62] – [57] [63] [58] ANT2 – Other mitochondria-non-ETC F13G3.7 (mitochondrial carrier protein) [13] ––––carrier – proteins C33F10.12 (mitochondrial phosphate [62] ––––phosphate – carrier protein) carriers K01C8.7 (mitochondrial FAD carrier protein) [13] [56, 57] – [58] – SLC25A32 – F54H12.1 (aco-2, aconitase) [62] – [57] [56] [56–58] ACO2 – ZK637.9 (tpk-1, thimaine pyrophosphokinase) [76] ––––TPK1 d) F43G9.1 (isocitrate dehydrogenase, [62] [57] [61] [58, 63] [57, 58, 61] IDH3A – alpha subunit) F59B8.2 (NADP-dependent [62] ––––IDH2 – isocitrate dehydrogenase) C37H5.8 (hsp-6, mortalin, mot-2/mtHSP70) c) c) [57, 63], c) [56] [56–58] mtHSP70 d)

a) Age: long-lived, Gro: slow larval growth, Lva: larval arrest, Ste/Fer: sterile or reduced fertility, Emb: embryonic lethal. b) HMAD: Human mitochondrial-associated disorder. c) This study. d) Potentially.

Biotechnol. J. 2007, 2, 584–595 www.biotechnology-journal.com

presentation of these associated phenotypes appears to 2.3 Consecutive generational RNAi be dependent upon both the type and degree of target disruption. Several of these associated phenotypes are more For studies involving frh-1 and nuo-2, N2 animals were easily scored than lifespan in worms. We believe that the cultured on pL4440-containing, nuo-2 RNAi-producing or coordinate appearance of these phenotypes is due to spe- frh-1 RNAi-producing HT115 (DE3) bacteria for one to cific compensatory pathways activated upon mitochon- three consecutive generations as described previously drial dysfunction. Loss of cellular compensation inevitably [15]. leads to the coordinate disappearance of these phenotypes, thus providing us with surrogate markers for mito- 2.4 Lifespan analysis chondrial dysfunction. In this study, we will describe our recent efforts show- Lifespan was recorded at 20°C using synchronous popu- ing how the C. elegans Mit mutants can be used as a nov- lations of 60–80 animals per RNAi dilution or generation. el and potentially powerful model to study neurodegener- Lifespan analyses began from the start of the egg laying. ative and other mitochondrial-related disorders in hu- Data were analyzed using the Log Rank Test as previous- mans. By identifying signaling pathways regulating the ly described [17]. various Mit phenotypes we hope to uncover similar signaling pathways that may be of relevance to offsetting hu- 2.5 Western blot analysis man mitochondrial-related pathologies. Whole-worm lysates were prepared from 200 animals in boiling 5% SDS, 0.02% β-mercaptoethanol, and 1 mM pro- 2 Materials and methods tease inhibitor cocktail (SIGMA P2714). Following protein quantitation (BCA, Pierce), protein levels were measured 2.1 Nematode maintenance by Western analysis, quantified using densitometry, then normalized against actin. Antibodies employed were Standard nematode culturing techniques were employed α-NUO-2 (MS112, MitoSciences, Oregon, 1:2000), α-actin [16]. The following strains were utilized: N2 (wild type), (AC-15, Sigma, 1:5000), α-ANT (MSA02, MitoSciences, SJ4100 [hsp-6::GFP(zcIs13)], DA465 [eat-2(ad465)], LG100 1:1000), α -complex IV, subunit I (MS404, MitoSciences, [sir-2.1(geIn3)], VC389 [frh-1(ok610)/mIn1[mIs14 dpy-10 1:1000), α-complex V, alpha subunit (MS507, Mito- (e128)]. Sciences, 1:1000).

2.2 Feeding RNAi 2.6 Microscopy

The following RNAi feeding constructs were derived from Individual worms were randomly selected from RNAi the Ahringer C. elegans RNAi library: hsp-6 (C37H5.8), plates and then arranged for image analysis on 2% atp-3 (F27C1.7), nuo-2 (T10E9.7), cco-1 (F26E4.9), cyc-1 agarose pads. Nomarski and fluorescence images were (C54G4.8), R53.4, D2030.4, tag-55 (F42A8.2), F26E4.6, collected using a PCO SensiCam charge-coupled diode F13G3.7, B0261.4, and W09C5.8. Feeding RNAi con- camera connected to a Zeiss Axioskop running Slide- structs against frh-1 and isp-1 have been described pre- Maker 4.0 software. All other images were collected using viously ([15] and Shane L. Rea, Nastascia Ventura and a digital camera connected to a dissecting microscope. Thomas E. Johnson, manuscript submitted, respectively). For RNAi dilutions, cultures of HT115 (DE3) containing empty vector (pL4440) or each gene of interest were pre- 3 Results and discussion pared in LB broth containing 100 μg/mL ampicillin and μ 5 g/mL tetracycline and grown overnight at 37°C. OD590 3.1 Mimicking human mitochondrial-associated diseases values were adjusted to 0.9 and bacteria mixed in the fol- in C. elegans lowing target gene to pL4440 ratio – 0:1, 1:10, and 1:0. For each mix, 250 μL was added to a 6 cm RNAi plate (NGM Most of the mitochondrial-related disorders in humans agar + 1 mM IPTG, 100 μg/mL ampicillin and 5 μg/mL are phenotypically complex; that is, different mitochon- tetracycline) and grown overnight at 23°C. Eighty N2 drial mutations can produce the same phenotype while eggs from a 4-h limited-lay were transferred from stan- the same mutation can lead to the appearance of different dard OP50 culture plates to each type of RNAi plate and phenotypes. Generally, such diseases have a chronic pro- allowed to develop at 20°C. gressive course and they present with symptoms of neurodegeneration, myopathy, and/or cardiomyopathy (encephalocardiomyopathies). The type and severity of a mitochondrial-related genetic lesion, its tissue-specific distribution as well as the activation of different compen-

Biotechnology Journal Biotechnol. J. 2007, 2, 584–595

satory pathways aimed at countering the lesion, dictate, in the end, the final presenting phenotype in patients. Similarly in C. elegans, disruption of different mitochondrial genes can produce the same phenotype, while disruptions in the same gene can result in the appearance of distinct phenotypic features depending on the degree of gene disruption (Table 1 and Shane L. Rea, Nastascia Ventura and Thomas E. Johnson, manuscript submitted). In an effort to establish novel models to study human mitochondrial-related diseases, we have used bacterial feeding RNAi to systematically disrupt the function of multiple genes in C. elegans that are orthologous to known genes involved in human mitochondrial pathologies. To mimic the chronic course of these diseases we have developed a simple but powerful technique to modulate the degree of inactivation of C. elegans mitochondrial proteins (Materials and methods, Shane L. Rea, Nas- tascia Ventura and Thomas E. Johnson, manuscript submitted). By increasing the potency of each bacterial feeding RNAi (either by using bacterial dilution or by crossgenerational RNAi feeding), we have observed that disruption of many mitochondrial genes leads to a classic Mit phenotype – that is, at low levels of RNAi-mediated disruption larval development is slowed, adult size and fertility are reduced, and lifespan is extended. Following a more severe disruption (either by increasing RNAi potency or by gene knockout (KO)), deleterious phenotypes are often aroused – larval development becomes arrested or animals become sterile and/or lifespan shortens pathologically. Interestingly, we found that the disruption of several target genes displayed no phenotypic conse- quence, while others showed only mild effects. We suspect that differences in the efficacy of RNAi-mediated target knockdown, coupled with some unique properties of C. elegans physiology (such as the persistence of maternally contributed products), likely explain much of the gross variability in phenotypes that we observed. Nonetheless, several new models of human diseases have been established, three of which we will briefly mention. HSP-6 is the C. elegans homolog of the mammalian mitochondrial protein HSP-70/mortalin/mthsp70 [18]. This protein, in cooperation with several other chaper- onins, functions in promoting efficient mitochondrial protein import and folding. Previous work has shown that alterations in the mitochondrial import machinery can lead to human disease [19, 20]. Moreover, several reports have

indicated that cells from patients affected by any one of a Figure 1. Phenotypes associated with hsp-6, nuo-2, and frh-1. Disruption variety of mitochondrial diseases express altered levels of in C. elegans. HT115 bacteria expressing dsRNA targeting hsp-6 (A and B), mRNAs and/or proteins encoding HSP-70 and other fac- nuo-2 (C and D), or frh-1 (E and F) were fed to N2 (wild-type) animals in tors belonging to the import machinery [21, 22]. When we undiluted form, or as a 1:10 mix with control bacteria (hsp-6 only), from fed wild type C. elegans with a 1/10 dilution of RNAi di- the time of hatching for one (hsp-6, nuo-2, P0), two (nuo-2, G1), or three rected against hsp-6, we observed that the larval devel- (frh-1) consecutive generations. frh-1 KO animals, and the parental wild- opment slowed and lifespan increased. When animals type behaving strain VC389, were cultured on OP50 bacteria. C. elegans lifespan is increased by ∼3 days when grown on this bacteria, relative to were fed undiluted hsp-6 RNAi they instead arrested as growth on HT115. The effect on size (A, C, E) and lifespan (B, D, F) was early larvae and lifespan was brought back to untreated- subsequently recorded. Animals were photographed 2 days (A), 8 days control levels (Figs. 1A and B). (C, P0), 5 days (C, G1), or 4 days (E) after hatching.

Biotechnol. J. 2007, 2, 584–595 www.biotechnology-journal.com

Complex I (NADH:ubiquinone oxidoreductase), of the ly dictate the final phenotypic mix seen at the organismal canonical ETC, is comprised of multiple protein subunits. level. Alternatively, different compensatory pathways In humans, disruptions to any of the 14 core complex I may be activated following different kinds of mitochondr- subunits results in the appearance of either Leigh’s syn- ial disruption and such pathways may be tissue specific. drome, LHON disease, or cardiomyopathy [23]. In worms, The Mit mutants provide an entrée into the elucidation of RNAi-mediated reduction in different complex I compo- such processes. nents leads to different phenotypes: Diminished expression of seven complex I genes (nuo-1 to nuo-5, D2030.4, 3.2 Pathways potentially regulating Mit mutant phenotypes and T20H4.5) has been previously reported to result in lifespan extension (Table 1). Loss of gas-1, on the other 3.2.1 A mitochondrial unfolded protein response hand, is pathologically life-shortening [24]. A single case Understanding which mitochondrial parameters and report exists of a patient with a compound mutation in which signaling pathways are affected following different nuo-2 leading to Leigh’s syndrome [23]. When we rein- levels of mitochondrial dysfunction in the Mit mutants vestigated the disruption of nuo-2 in C. elegans using our could be extremely relevant to shedding light on different feeding RNAi dilution methodology we found that the phases of disease progression in humans. Several features lifespan was greatly extended even using undiluted are generally altered in response to mitochondrial dys- RNAi, as previously reported [14]. When nuo-2 levels were function including changes in protein folding, membrane more severely reduced following a crossgenerational import, free radical production, ATP levels, and nucleotide feeding RNAi strategy, animals displayed an even greater pools [2, 3]. Ron and coworkers [25] reported that disrup- life extension, yet larval development became arrested at tion of normal mitochondrial function by RNAi directed the early L3 stage (Figs. 1C and D). Larval-arrest associat- against a variety of mitochondrial proteins (including, ed with life-extension may therefore reflect differences in paradoxically, the mitochondrial heatshock protein HSP- the energy threshold for the progression of development 6), or by chemical inhibition using ethidium bromide, in- versus cellular maintenance. duced a mitochondrial-specific, unfolded protein re-

Complexes I, II, and III of the mitochondrial ETC as sponse (UPRmito) [25]. This response was detectable using well as aconitase of the tricarboxylic acid cycle, each con- a hsp-6promoter::GFP reporter construct. tain Fe-S clusters. Synthesis of these components re- Using the same hsp-6promoter::GFP reporter system, we quires the nuclear-encoded mitochondrial protein fratax- have observed that several Mit mutants constitutively ac-

in that, as mentioned above, leads to FA when disrupted tivate a UPRmito response (Fig. 2 and Table 2). This rein humans. In worms, mild reduction of frh-1 by RNAi sponse was not universal, however, since not all the Mit leads to life extension. When reduced more severely, us- mutants activated the reporter. Furthermore, we found

ing a crossgenerational feeding strategy on an RNAi en- that the level of hsp-6promoter::GFP activation in those Mit hancing strain, lifespan shortens (Shane L. Rea, Nastascia mutants that did turn it on was dependent upon the de- Ventura and Thomas E. Johnson, manuscript submitted gree of target gene inactivation. For most of the Mit mu- and Figs. 1E and F). One interesting observation that we tants, higher levels of target disruption led to higher lev- made involved animals genetically ablated for frh-1. els of reporter induction but this did not necessarily cor- These animals would seemingly represent the severest relate with greater lifespan increases (Fig. 2, Table 2 and level of frh-1 knockdown. However, we observed that de- data not shown). These findings suggest that there may spite arresting as early larvae, they are still long lived. This well be an optimal window of hsp-6 expression that cor- finding illustrates an important point about worm biology relates with maximal lifespan increase and that very – that maternally-contributed product can sometimes strong reporter induction presumably instead reflects ma- mask a genetic phenotype. KO animals were derived from jor mitochondrial problems. Consistent with the notion of a heterozygous hermaphoditic parent and presumably ei- an optimal window of hsp-6 expression leading to life ex- ther frh-1 mRNA or protein was passed onto the homozy- tension, Yokoyama et al. [26] have reported that overex- gous frh-1 animals, confounding the longevity data. pression of hsp-6 alone can increase C. elegans lifespan.

Our present results reinforce our earlier findings We conclude therefore that hsp-6promoter::GFP induction in (Shane L. Rea, Nastascia Ventura and Thomas E. John- those Mit mutants that turn it on likely reflects a general son, manuscript submitted), showing that titrating RNAi mitochondrial stress response and, although probably against different mitochondrial components can result in necessary for their survival, it may not be sufficient to ac- seemingly vastly different phenotypic outcomes. Further- count entirely for their life extension. more, the precise degree of target inactivation appears to Isp-1(qm150) is one of the few genetically defined C. play a major role in determining the phenotypic outcome. elegans Mit mutants. Isp-1 encodes the Rieske iron–sul- The core Mit phenotypes appear to be differentially sen- fur protein subunit of complex III and the qm150 allele sitive to mitochondrial disruption. Different target tissues contains a missense mutation, most likely affecting the control each phenotype so we suspect that variations in redox potential of the protein. This mutant is character- the efficacy of target knockdown in each tissue ultimate- ized by increased longevity, reduced fertility, and delayed

Biotechnology Journal Biotechnol. J. 2007, 2, 584–595

notion that different stress response pathways are differentially activated following mitochondrial dysfunction and that they are dependent upon both the kind and severity of gene inactivation.

3.2.2 Calorie restriction As in other species, in C. elegans caloric restriction delays aging [28, 29]. How exactly organisms benefit from this in- tervention is unclear, but it has been proposed that a mild metabolic stress could cause cells to respond by modulat- ing their energy metabolism, redox status, protein biosynthesis, mitochondrial function, and/or genomic integrity. Ultimately, it is thought that there is an enhanced turnover of damaged molecules and/or the prevention of their formation, leading to improved cellular and organismal survival [30–32]. We believe that a related phenome- non may, in part, also be functioning in the Mit mutants where an initial mild mitochondrial stress could similarly promote the induction of a whole organism’s protective response. Support for this idea comes from the observations of Tsang and Lemire [33] who noticed that mitochondrial mutants that arrested as early larvae have empty guts and appear starved. Two interventions which increase C. elegans lifespan with a caloric restriction-like mechanism are genetic mutations in the eat genes (which produce a defect in pha- ryngeal pumping) [34], and overexpression of sir-2.1(+) Figure 2. Unfolded protein response in the Mit mutants. Development is (which encodes an NAD-dependent deacetylase that delayed, and a mitochondrial-specific UPRmito is activated, in multiple is activated following conditions of food deprivation) RNAi-induced Mit mutants. N2 animals were exposed to RNAi against [35–37]. A powerful approach afforded by the C. elegans 1. 2. various mitochondrial components: left to right: control (vector), hsp- system is the ability to map genes to known or novel sig- 6, 3. nuo-2 (complex I), 4. D2030.4 (complex I), 5. tag-55 (complex II), 6. isp-1 (complex III), 7. cco-1 (complex IV), 8. W09C5.8 (complex IV), naling pathways using epistatic analysis. Using such an 9. cyc-1 (complex IV), 10. F26E4.6 (complex IV), 11. atp-3 (complex V), approach, we have found that RNAi directed against 12. R53.4 (complex V), 13. frh-1, 14. F13G3.7 (carrier protein), 15. B0261.4 frataxin further prolongs the longevity of both eat-2 loss of (rRNA) from the time of hatching (except frh-1 which had been cultured function (Fig. 3A) and sir-2.1(+) overexpressing strains on RNAi for two prior generations), then imaged as one day-old adults us- (Fig. 3B). This synergistic effect suggests two opposite in- ing normaski (A, B top panels), and fluorescence (A, B, bottom panels) terpretations which will need a further investigation: optics. All the RNAi targets except tag-55 have been shown previously to Frataxin might define a completely different longevity result in a bona fide Mit phenotype. For many of the targeted genes, the effect of RNAi on reporter induction is concentration dependent (compare pathway than either that specified by sir-2.1(+) and eat-2; or in fact it might work on the same pathway, reinforcing 1:10 RNAi dilution [A], with undiluted RNAi [B]). UPRmito was detected using a hsp-6promoter::GFP reporter strain [25]. the function of both alterations.

3.2.3 Exploring mitochondrial biogenesis in the Mit mutants embryonic and postembryonic development [27]. When The beneficial effects of caloric restriction have been as-

we crossed the hsp-6promoter::GFP reporter strain with isp- cribed to the induction of mitochondrial biogenesis and to 1(qm150) we found that the GFP reporter was robustly ac- the improvement of bioenergetic efficiency [38]. Peroxi- tivated. Identical results were also found using isp-1 RNAi some proliferative-activated receptor coactivator 1

in the hsp-6promoter::GFP reporter background (Fig. 2B). We (PGC1) has emerged as a central regulator in mammalian are currently using our isp-1(qm150); hsp-6promoter::GFP cells of both the adaptive response to caloric deprivation strain to look for both genes and drugs that can modulate [39] and mitochondrial biogenesis [40]. Many of the mito- expression of the Isp-1 phenotype as well as the hsp- chondrial and energy metabolism genes that are under

6promoter::GFP reporter. Similarly, we are also employing the control of PGC1 in mammals are also repressed with other GFP reporter strains that are indicative of a variety aging. This repression can be partially reversed by caloric of additional stress responses (gst-4::GFP, etc.), to eluci- restriction [41–43]. Several metabolites are altered in cells date genes and pathways affected in the isp-1 mutant following the mitochondrial dysfunction including ROS, background. Our preliminary observations support the Ca2+, NADH, and AMP. Changes in these compounds can

Biotechnol. J. 2007, 2, 584–595 www.biotechnology-journal.com e) rage level of GFP cluster protein synthesis 788423 e) 8866625 el. e) pective ETC complex. the highest). Reporter expression in vector control treatment (background) was defined 27622577323 38825 = 40 worms/RNAi, 20°C) and then GFP expression was scored at +48 h and then again at +72 h by two different investigators. The ave n Larval stages; YA: Young Adult; GA: Gravid Adult. 123456789101112131415 – Chaperone I I II III IV IV IV IV V VrRNA Fe-S Carrier 287738726788333 28885885 GA L3 arrest GA GA GA L4/YA YA/GA GA GA GA L3 arrest L4 GA GA GA C. elegans d) d) a) expression level in different Mit mutant backgrounds c) c) ::GFP promoter ::GFP expression level ::GFP expression level b) Hsp-6 promoter promoter expression in each population was subjectively assigned a value from 1 to 8 (with being the lowest level of and as 2. 48 h e) Heterogeneous GFP expression across population. Some worms show this level of expression, others only background lev Table 2. RNAi targetMitochondrial Function Fig. 2 worm Developmental stage ControlUndiluted RNAi hsp-61:10 RNAi nuo-2hsp-6 D2030.4 tag-55Undiluted RNAi isp-11:10 RNAi cco-1 L4/YA W09C5.8Developmental stage cyc-1 L3Undiluted RNAi F26E4.6 L4/YA1:10 RNAi atp-3 L4/YA L3/L4 R53.4hsp-6 L4/YA YA frh-1Undiluted RNAi L4 F13G3.71:10 RNAi B0261.4 2 YAa) Roman numerals refer to complexes I, II, III, IV, and V of the ETC. Genes with these annotations encode components res L3 L4 late L3 8 L4/YA GA L4 L4 6 L4/YA GA L4/YA 4 late L3 L4 GA 2 late L3 L4 L4 GA 8 YA GA L4 7 L3-YA L3-YA GA L4/YA L4/YA 5 GA L4/YA GA YA GA GA GA GA GA GA GA 72 h b) Representative fluorescent images of 72 h worms are shown in Fig. 2. c) L1, L2, L3, and L4 represent consecutive d) N2 animals were exposed to RNAi from the time of hatching (

Biotechnology Journal Biotechnol. J. 2007, 2, 584–595

panded subsarcolemmal mitochondria, are common to myopathies associated with several HMADs and presumably represent an attempt by cells to counter their reduced ETC activity [45]. Mitochondrial biogenesis may also then be one of the strategies employed by the Mit mutants to counter their mitochondrial defects. Interest- ingly, in our first attempts to characterize the mitochondrial morphology at the ultrastructural level using electron microscopy (EM) in a C. elegans Mit mutant, we have observed that compared to control animals, frh-1 RNAi- treated animals contain an increased number of mitochondria and of thin membranous bridges connecting them which may be indicative of mitochondrial biogenesis (Gin Fonte, N.V., unpublished observation). Other studies also support the notion that mitochondrial dy- namics are dysregulated in the Mit mutants [13]. As an alternative approach to studying mitochondrial biogenesis in the Mit mutants we have focused our at- tention on changes in mitochondrial protein expression. mtDNA copy number increases during C. elegans development, mostly in coincidence with germline maturation [48]. C. elegans embryos have about 2.5 × 104 copies of mtDNA and this number remains unchanged until the L3 larval stage. By the beginning of the L4 larval stage this Figure 3. Epistatic analysis of life extension pathways in the frh-1 Mit mu- value has increased fivefold, 50% of this increase is at- tant. N2 wild type (A and B, circles), eat-2(ad465) (A, triangles), and sir-2.1 tributable to somatic cells duplication while 50% to the over-expressing animals (B, triangles) were subjected to vector control germline expansion (by both oocytes and sperm). mDNA (closed symbols) or frataxin RNAi (open symbols) for three consecutive copy number then increases a further sixfold by the time generations and lifespan recorded in the third generation. Eat-2(ad465) young adulthood is reached, and this latter expansion is and sir-2.1 over-expressing strains are both long-lived. Exposure to frh-1 due entirely to hermaphroditic oocyte production. These RNAi extends the lifespan of both strains even further. increases in mtDNA copy number during nematode development suggest a parallel increase in the number of mitochondria, or at least of mitochondrial protein expres- be sensed by several proteins, such as AMP kinase, sion, although so far this has not been formally shown. p38/JNK stress/mitogen-activated kinase, mammalian With the few antibodies available against C. elegans mi- target of rapamicin (mTOR), the deacetylase sirtuin 1 tochondrial proteins, we have found that at least one nu- (SIRT1), and calcium/calmodulin-dependent protein ki- clear-encoded subunit of complex I (NUO-2), and a mito- nase IV (CaMKIV) [39], all of which finely tune PGC1 acti- chondrial-encoded subunit of complex IV (COX1), both vation. Stimulation of PGC1 induces the doubling of clearly increase during the development, in accord with mtDNA [44] as well as a plethora of transcription factors mitochondrial expansion. Surprisingly, the α-subunit of (PPARγ, NRF1, NRF2, and mtTFA) that in turn regulate a complex V seemed to remain mostly unchanged (Fig. 4A). suite of genes (such as COXIV, COXI, b-ATP synthetase, Since we analyzed whole worm extracts, we presume that cyt-c) necessary for the proper assembly and integration these differential changes reflect tissue-specific alter- of new mitochondria into the existing mitochondrial retic- ations. Expansion of the gonad presumably underlies ulum [45]. The C. elegans genome apparently lacks or- many of these differences. Alternatively, such changes thologs of the PGC1 proteins [46]. However, a partial func- could reflect larval-specific subunit requirements during tional overlap between mammalian PGC1 and C. elegans animal development. MDT-15 has been strongly suggested by the data of When we explored the expression of our small selec- Taubert et al. [47]. Interestingly, we recently identified a tion of mitochondrial proteins in three RNAi-induced Mit target of the MDT-15 coactivator in a screen we undertook mutants (frh-1, cco-1, and nuo-2), we invariably found that for factors affecting isp-1 viability (unpublished observa- multiple complexes were affected by the loss of a single tions). target protein (Fig. 4B). One interpretation is that reduc- When a metabolic stress (such as cold or exercise) tion of multiple complexes merely reflects the loss of go- places chronic demands upon ATP level, many cells re- nadal cells, since each RNAi also affects fertility. If this spond by altering their metabolism and increasing mito- were true, one would have expected to have seen a con- chondrial biogenesis. Ragged-red fibers, with their ex- sistent decrease in all the ETC subunits, but, as shown in

Biotechnol. J. 2007, 2, 584–595 www.biotechnology-journal.com

Fig. 4B, this did not appear to be the case. While loss of mitochondrial-encoded COX1 did consistently decrease following each RNAi treatment, and probably reflecting, in part, reduced gonad expansion, the other ETC subunits we analyzed did not follow the same trend. Indeed, in some instances the other ETC subunits actually increased in concentration. This effect becomes more apparent when the data of Fig. 4B are reorganized into the format of Fig. 4C. The unique pattern of subunit disruption following each RNAi treatment is obvious. Our current findings on subunit alterations are consistent with earlier studies in mammalian cells showing that different mitochondrial mutations lead to distinct changes in mitochondrial ETC protein expression profiles [49, 50]. Such findings support the relevance of the Mit mutant model in the study of human mitochondrial-related diseases. Presently, we are establishing global profiles of mitochondrial protein changes, across different levels of gene inactivation, for a variety of Mit mutants. It is an- ticipated that such data will help shed light on different phases of the corresponding human diseases.

4 Concluding remarks

Many human neurodegenerative disorders are associated with mitochondrial dysfunction and severely compro- mised energy generation, and arise because of genetic defects in the mitochondrial or nuclear genome [51]. Sev- eral of the affected genes have orthologs in C. elegans. We have successfully disrupted a number of these genes in worms using RNAi in an effort to establish new models to study the human disorders. In this paper, we have out- lined our progress on this front. Aging is the major risk factor for the appearance of neurodegenerative disorders and alteration in mitochondrial function proceeds, or is associated with, several of Figure 4. Mitochondrial protein composition is differentially altered in these diseases. The kinds of mitochondrial alterations several Mit mutants. (A) Western analysis showing the expression profile that are associated with neurodegeneration resemble of four mitochondrial proteins (complex I (NDUSF3 subunit), complex IV (COX1 subunit), complex V (α subunit), ANT (adenine nucleotide trans- those that occur normally during the aging process and porter)) during the postembryonic development of C. elegans. COX1 is mi- caloric restriction has been shown to delay both aging tochondrially-encoded. Actin is shown for normalization purposes and all and age-associated diseases. It is probable, then, that the data are representative of three independent experiments. (L1–L4, lar- there is a significant overlap in the pathogenic mecha- val stages one through four; GA, gravid adult; 7do, 7 day-old). (B and C) nisms leading to both processes. N2 animals were exposed to a 1:10 dilution of RNAi against cco-1 or nuo-2 Worms appear to be able to compensate for the partial from the time of hatching, or to undiluted frh-1 RNAi for three consecutive mitochondrial dysfunction by paradoxically increasing generations, and then all processed as 1 day-old gravid adults for Western analysis against the same four proteins described in (A). Protein levels their lifespan. We are currently using this phenotype as were densitometrically quantitated using Image J (NIH). Sample loading well as others, to gain insight into the ways mutant was normalized on the basis of actin. Subunit alterations are expressed in worms offset the severity of their mitochondrial disrup- fold change relative to vector-treatment alone (pl4440). Error bars repre- tions with the ultimate hope of applying this knowledge sent SEM of the fold change from experiments carried out on two different to the prevention of the corresponding human diseases. biological samples. (B) and (C) are alternative representations of the Recent advances have permitted C. elegans to be used as same data. a test bed for drug screening [52, 53]. The pharmacologic alteration of the Mit phenotype represents a promising strategy to discover new targets potentially exploitable for

Biotechnology Journal Biotechnol. J. 2007, 2, 584–595

early preventive therapy in human neurodegenerative [18] Heschl, M. F., Baillie, D. L., Characterization of the hsp70 multigene diseases and other mitochondrial-related disorders. family of Caenorhabditis elegans. DNA 1989, 8, 233–243. [19] Koehler, C. M., Leuenberger, D., Merchant, S., Renold, A. et al., Hu- man deafness dystonia syndrome is a mitochondrial disease. Proc. Natl. Acad. Sci. USA 1999, 96, 2141–2146. The authors wish to acknowledge Professor Thomas E. [20] Huckriede, A., Agsteribbe, E., Decreased synthesis and inefficient Johnson for kindly providing laboratory space and mitochondrial import of hsp60 in a patient with a mitochondrial en- reagents; Dr. Christopher Link and Ms. Virginia Fonte for cephalomyopathy. Biochim. Biophys. Acta 1994, 1227, 200–206. permission to cite unpublished EM data and Professor [21] Rungi, A. A., Primeau, A., Nunes Christie, L., Gordon, J. W. et al., Roberto Testi; Alison Kell (B.Sc.) and Ben Gurney (B.Sc.) Events upstream of mitochondrial protein import limit the oxidative for technical assistance; and the Caenorhabditis Genetics capacity of fibroblasts in multiple mitochondrial disease. Biochim. Centre (CGC) for provision of strains. Financial support Biophys. Acta. 2002, 1586, 146–154. [22] Joseph, A. M., Rungi, A. A., Robinson, B. H., Hood, D. A., Compen- was provided by the National Institute on Aging (NIA) to satory responses of protein import and transcription factor expres- S.L.R.; by the National Ataxia Foundation (NAF) and the sion in mitochondrial DNA defects. Am. J. Physiol. Cell. Physiol. Italian Association on Cancer Research (AIRC) to N.V. 2004, 286, C867–C875. [23] Benit, P., Slama, A., Cartault, F. Giurgea, I. et al., Mutant NDUFS3 subunit of mitochondrial complex I causes Leigh syndrome. J. Med. 5 References Genet. 2004, 41, 14–17. [24] Kayser, E. B., Morgan, P. G., Hoppel, C. L., Sedensky, M. M., Mito- chondrial expression and function of GAS-1 in Caenorhabditis ele- [1] Kwong, J. Q., Beal, M. F., Manfredi, G., The role of mitochondria in gans. J. Biol. Chem. 2001, 276, 20551–20558. inherited neurodegenerative diseases. J. Neurochem. 2006, 97, [25] Yoneda, T., Benedetti, C., Urano, F., Clark, S. G. et al., Compartment- 1659–1675. specific perturbation of protein handling activates genes encoding [2] Wallace, D. C., Mitochondrial diseases in man and mouse. Science mitochondrial chaperones. J. Cell. Sci. 2004, 117, 4055–4066. 1999, 283, 1482–1488. [26] Yokoyama, K., Fukumoto, K., Murakami, T., Harada, S. et al., Ex- [3] Singh, K. K., Mitochondria damage checkpoint, aging, and cancer. tended longevity of Caenorhabditis elegans by knocking in extra Ann. N. Y. Acad. Sci. 2006, 1067, 182–190. copies of hsp70F, a homolog of mot-2 (mortalin)/mthsp70/Grp75. [4] Lin, M. T., Beal, M. F., Mitochondrial dysfunction and oxidative FEBS Lett. 2002, 516, 53–57. stress in neurodegenerative diseases. Nature 2006, 443, 787–795. [27] Feng, J., Bussiere, F., Hekimi, S., Mitochondrial electron transport is [5] Bulteau, A. L., Ikeda-Saito, M., Szweda, L. I., Redox-dependent mod- a key determinant of life span in Caenorhabditis elegans. Dev. Cell. ulation of aconitase activity in intact mitochondria. Biochemistry 2001, 1, 633–644. 2003, 42, 14846–14855. [28] Houthoofd, K., Gems, D., Johnson, T. E., Vanfleteren, J. R., Dietary re- [6] Karlberg, T., Schagerlof, U., Gakh, O., Park, S. et al., The structures striction in the nematode Caenorhabditis elegans. Interdiscip. Top. of frataxin oligomers reveal the mechanism for the delivery and Gerontol. 2007, 35, 98–114. detoxification of iron. Structure 2006, 14, 1535–1546. [29] Longo, V. D., Finch, C. E., Evolutionary medicine: From dwarf mod- [7] Rossignol, R., Faustin, B., Rocher, C., Malgat, M. et al., Mitochondrial el systems to healthy centenarians?. Science 2003, 299, 1342–1346. threshold effects. Biochem. J. 2003, 370, 751–762. [30] Yu, Z., Luo, H., Fu, W., Mattson, M. P., The endoplasmic reticulum [8] Ventura, N., Rea, S. L., Testi, R., Long-lived, C. elegans Mitochon- stress-responsive protein GRP78 protects neurons against excito- drial mutants as a model for human mitochondrial-associated dis- toxicity and apoptosis: Suppression of oxidative stress and stabi- eases. Exp. Gerontol. 2006, 41, 974–991. lization of calcium homeostasis. Exp. Neurol. 1999, 155, 302–314. [9] Coppola, G., Choi, S. H., Santos, M. M., Miranda, C. J. et al., Gene [31] Lee, C. K., Weindruch, R., Prolla, T. A., Gene-expression profile of the expression profiling in frataxin deficient mice: Microarray evidence ageing brain in mice. Nat. Genet. 2000, 25, 294–297. for significant expression changes without detectable neurodegen- [32] Tavernarakis, N., Driscoll, M., Caloric restriction and lifespan: A role eration. Neurobiol. Dis. 2006, 22, 302–311. for protein turnover?. Mech. Ageing Dev. 2002, 123, 215–229. [10] Harman, D., Aging: A theory based on free radical and radiation [33] Tsang, W. Y., Lemire, B. D., The role of mitochondria in the life of the chemistry. J. Gerontol. 1956, 11, 298–300. nematode, Caenorhabditis elegans. Biochim. Biophys. Acta. 2003, [11] Balaban, R. S., Nemoto, S., Finkel, T., Mitochondria, oxidants, and 1638, 91–105. aging. 2005, Cell 120, 483–495. [34] Lakowski, B., Hekimi, S., The genetics of caloric restriction in [12] Beal, M. F., Mitochondria take center stage in aging and neurode- Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 1998. 95, generation. Ann. Neurol. 2005, 58, 495–505. 13091–13096. [13] Lee, S. S., Lee, R. Y., Fraser, A. G., Kamath, R. S. et al., A systematic [35] Tissenbaum, H. A., Guarente, L., Increased dosage of a sir-2 gene ex- RNAi screen identifies a critical role for mitochondria in C. elegans tends lifespan in Caenorhabditis elegans. Nature 2001, 410, longevity. Nat. Genet. 2003, 33, 40–48. 227–230. [14] Dillin, A., Hsu, A. L., Arantes-Oliveira, N., Lehrer-Graiwer, J. et al., [36] Wood, J. G., Rogina, B., Lavu, S., Howitz, K. et al., Sirtuin activators Rates of behavior and aging specified by mitochondrial function mimic caloric restriction and delay ageing in metazoans. Nature during development. Science 2002, 298, 2398–2401. 2004, 430, 686–689. [15] Ventura, N., Rea, S., Henderson, S. T., Condo, I. et al., Reduced ex- [37] Wang, Y., Tissenbaum, H. A., Overlapping and distinct functions for pression of frataxin extends the lifespan of Caenorhabditis elegans. a Caenorhabditis elegans SIR2 and DAF-16/FOXO. Mech. Ageing Aging Cell 2005, 4, 109–112. Dev. 2006, 127, 48–56. [16] Wood, W. B. (Ed.), The Nematode Caenorhabditis elegans. Cold [38] Lopez-Lluch, G., Hunt, N., Jones, B., Zhu, M. et al., Calorie restric- Spring Harbor Laboratory, New York 1988. tion induces mitochondrial biogenesis and bioenergetic efficiency. [17] Johnson, T. E., Wood, W. B., Genetic Analysis of Life-Span in Proc. Natl. Acad. Sci. USA 2006, 103, 1768–1773. Caenorhabditis elegans. PNAS 1982, 79, 6603–6607.

Biotechnol. J. 2007, 2, 584–595 www.biotechnology-journal.com

[39] Rodgers, J. T., Lerin, C., Haas, W., Gygi, S. P. et al., Nutrient control [58] Sonnichsen, B., Koski, L. B., Walsh, A., Marschall, P., et al., Full- of glucose homeostasis through a complex of PGC-1alpha and genome RNAi profiling of early embryogenesis in Caenorhabditis el- SIRT1. Nature 2005, 434, 113–118. egans. Nature 2005, 434, 462–469. [40] Wu, Z., Puigserver, P., Andersson, U., Zhang, C. et al., Mechanisms [59] Schuelke, M., Smeitink, J., Mariman, E., Loeffen, J. et al., Mutant controlling mitochondrial biogenesis and respiration through the NDUFV1 subunit of mitochondrial complex I causes leukodystrophy thermogenic coactivator PGC-1. Cell 1999, 98, 115–124. and myoclonic epilepsy. Nat. Genet. 1999, 21, 260–261. [41] Mootha, V. K., Lindgren, C. M., Eriksson, K. F., Subramanian, A. et [60] Hansen, M., Hsu, A. L., Dillin, A., Kenyon, C., New genes tied to en- al., PGC-1alpha-responsive genes involved in oxidative phosphory- docrine, metabolic, and dietary regulation of lifespan from a lation are coordinately downregulated in human diabetes. Nat. Caenorhabditis elegans genomic RNAi screen. PLoS Genet. 2005, 1, Genet. 2003, 34, 267–273. 119–128. [42] Patti, M. E., Butte, A. J., Crunkhorn, S., Cusi, K. et al., Coordinated [61] Fraser, A. G., Kamath, R. S., Zipperlen, P., Martinez-Campos, M., et reduction of genes of oxidative metabolism in humans with insulin al., Functional genomic analysis of C. elegans chromosome I by sys- resistance and diabetes: Potential role of PGC1 and NRF1. Proc. Natl. tematic RNA interference. Nature 2000, 408, 325–330. Acad. Sci. USA 2003, 100, 8466–8471. [62] Hamilton, B., Dong, Y., Shindo, M., Liu, W. et al., A systematic RNAi [43] Corton, J. C., Brown-Borg, H. M., Peroxisome proliferator-activated screen for longevity genes in C. elegans. Genes. Dev. 2005, 19, receptor gamma coactivator 1 in caloric restriction and other mod- 1544–1555. els of longevity. J. Gerontol. A Biol. Sci. Med. Sci. 2005, 60, 1494–509. [63] Rual, J. F., Ceron, J., Koreth, J., Hao, T. et al., Toward improving [44] Puigserver, P., Wu, Z., Park, C. W., Graves, R. et al., A cold-inducible Caenorhabditis elegans phenome mapping with an ORFeome- coactivator of nuclear receptors linked to adaptive thermogenesis. based RNAi library. Genome. Res. 2004, 14, 2162–2168. Cell 1998, 92, 829–839. [64] Benit, P., Chretien, D., Kadhom, N., de Lonlay-Debeney, P. et al., [45] Chabi, B., Adhihetty, P. J., Ljubicic, V., Hood, D. A., How is mito- Large-scale deletion and point mutations of the nuclear NDUFV1 chondrial biogenesis affected in mitochondrial disease?. Med. Sci. and NDUFS1 genes in mitochondrial complex I deficiency. Am. J. Sports Exerc. 2005, 37, 2102–2110. Hum. Genet 2001, 68, 1344–1352. [46] Knutti, D., Kaul, A., Kralli, A., A tissue-specific coactivator of steroid [65] Loeffen, J. L., Triepels, R. H., van den Heuvel, L. P. Schuelke, M. et receptors, identified in a functional genetic screen. Mol. Cell. Biol. al., cDNA of eight nuclear encoded subunits of NADH:ubiquinone 2000, 20, 2411–2422. oxidoreductase: Human complex I cDNA characterization complet- [47] Taubert, S., Van Gilst, M. R., Hansen, M., Yamamoto, K. R., A Medi- ed. Biochem. Biophys. Res. Commun. 1998, 253, 415–422. ator subunit, MDT-15, integrates regulation of fatty acid metabolism [66] Maeda, I., Kohara, Y., Yamamoto, M. Sugimoto, A., Large-scale by NHR-49-dependent and -independent pathways in C. elegans. analysis of gene function in Caenorhabditis elegans by high- Genes Dev. 2006, 20, 1137–1149. throughput RNAi. Curr. Biol. 2001, 11, 171–176. [48] Tsang, W. Y., Lemire, B. D., Mitochondrial genome content is regu- [67] Haut, S., Brivet, M., Touati, G., Rustin, P. et al., A deletion in the hu- lated during nematode development. Biochem. Biophys. Res. Com- man QP-C gene causes a complex III deficiency resulting in hypo- mun. 2002, 291, 8–16. glycaemia and lactic acidosis. Hum. Genet. 2003, 113, 118–122. [49] Marusich, M. F., Robinson, B. H., Taanman, J. W. Kim, S. J. et al., Ex- [68] Campuzano, V., Montermini, L., Molto, M. D., Pianese, L. et al., pression of mtDNA and nDNA encoded respiratory chain proteins in Friedreich’s ataxia: Autosomal recessive disease caused by an in- chemically and genetically-derived Rho0 human fibroblasts: A com- tronic GAA triplet repeat expansion. Science 1996, 271, 1423–1427. parison of subunit proteins in normal fibroblasts treated with ethid- [69] Fernandez, A. G., Gunsalus, K. C., Huang, J., Chuang, L. S. et al., ium bromide and fibroblasts from a patient with mtDNA depletion New genes with roles in the C. elegans embryo revealed using RNAi syndrome. Biochim. Biophys. Acta 1997, 1362, 145–159. of ovary-enriched ORFeome clones. Genome. Res. 2005, 15, [50] Li, K., Neufer, P. D., Williams, R. S., Nuclear responses to depletion of 250–259. mitochondrial DNA in human cells. Am. J. Physiol. 1995, 269, [70] Wimplinger, I., Morleo, M., Rosenberger, G., Iaconis, D. et al., Muta- C1265–C1270. tions of the mitochondrial holocytochrome c-type synthase in X- [51] Wallace, D. C., A mitochondrial paradigm of metabolic and degen- linked dominant microphthalmia with linear skin defects syndrome. erative diseases, aging, and cancer: A dawn for evolutionary medi- Am. J. Hum. Genet. 2006, 79, 878–889. cine. Annu. Rev. Genet. 2005, 39, 359–407. [71] Asencio, C., Rodriguez-Aguilera, J. C., Ruiz-Ferrer, M., Vela, J., [52] Kokel, D., Xue, D., A Class of Benzenoid Chemicals Suppresses Navas, P., Silencing of ubiquinone biosynthesis genes extends life Apoptosis in C. elegans. Chembiochem 2006, 7, 2010–2015. span in Caenorhabditis elegans. FASEB J. 2003, 17, 1135–1137. [53] Artal-Sanz, M., de Jong, L., Tavernarakis, N., Caenorhabditis ele- [72] Wong, A., Boutis, P., Hekimi, S., Mutations in the clk-1 gene of gans: A versatile platform for drug discovery. Biotechnol. J. 2006, 1, Caenorhabditis elegans affect developmental and behavioral timing. 1405–1418. Genetics 1995, 139, 1247–1259. [54] Tsang, W. Y., Sayles, L. C., Grad, L. I., Pilgrim, D. B., Lemire, B. D.,, [73] Lakowski, B., Hekimi, S., Determination of life-span in Caenorhab- Mitochondrial respiratory chain deficiency in Caenorhabditis ele- ditis elegans by four clock genes. Science 1996, 272, 1010–1013. gans results in developmental arrest and increased life span. J. Biol. [74] Kiss, H., Kedra, D., Yang, Y., Kost-Alimova, M. et al., A novel gene Chem. 2001, 276, 32240–32246. containing LIM domains (LIMD1) is located within the common [55] Hu, J., Barr, M. M., ATP-2 interacts with the PLAT domain of LOV-1 eliminated region 1 (C3CER1) in 3p21.3. Hum. Genet. 1999, 105, and is involved in Caenorhabditis elegans polycystin signaling. Mol. 552–559. Biol. Cell. 2005, 16, 458–469. [75] t Hart, L. M., Hansen, T., Rietveld, I., Dekker, J. M. et al., Evidence [56] Kamath, R. S., Fraser, A. G., Dong, Y., Poulin, G. et al., Systematic that the mitochondrial leucyl tRNA synthetase (LARS2) gene repre- functional analysis of the Caenorhabditis elegans genome using sents a novel type 2 diabetes susceptibility gene. Diabetes 2005, 54, RNAi. Nature 2003, 421, 231–237. 1892–1895. [57] Simmer, F., Moorman, C., van der Linden, A. M., Kuijk, E. et al., [76] de Jong, L., Meng, Y., Dent, J., Hekimi, S., Thiamine pyrophosphate Genome-wide RNAi of C. elegans using the hypersensitive rrf-3 biosynthesis and transport in the nematode Caenorhabditis elegans. strain reveals novel gene functions. PLoS Biol. 2003, 1, E12. Genetics 2004, 168, 845–854.